Reports are out that AMD plans on cutting prices by up to 50%. This is bad news for a company that just recently couldn't keep up with demand and was selling their chips at a premium. Unfortunaly for AMD that all changed when the Core2 was announced. This new architecture along with a price war has Intel positioned to take back the market share loss as well as put AMD back in the red. Even worse for AMD is the fact that they are now in much greater debt than they have ever been. AMD's response is at least six months away and for the tech industry that is like a lifetime. I cannot see the future, but what I can see is that AMD will be in a world of hurt for the next 6-12 months.
What you should realize about AMD by now is that they are quite used to operating at a loss. They survived their 1.2 billion annual loss in 2002 after all.
No, AMD will be around forever despite their usually poor financial performance. If things get too bad for them, they can always count on IBM, the German taxpayers, and their extremely loyal army of fanboys to come to their rescue.
As RichPLS stated, no company can operate at a loss... and, that's not even a matter of getting used to.
Plus, I don't think IBM, US gov., the AMD fan-base or their investors will be able/willing to support a moribund company; what would the point be in supporting a chip manufacturer which no-longer can produce high-standard, competitive chips?! moreover, I don't believe this to be the case, either.
AMD will be facing a hard time, as many have already referred & as conjunctural circumstances forecasted.
It's my conviction that, given the actual circumstances, AMD should opt for a somewhat different approach to
computing, through the use of off-chip/off-die co-processors. As a mere conjectural exercise, I don't think AMD will have a new microarchitecture soon enough to compete with Intel, in parallel; K8L could hardly be called a new uArch (as far as available data goes) and Intel's putting forward a two-by-two year uArch revision; Fab conversion to upgraded 65nm/300mm process manufacturing is too late, too expensive (AMD has made massive investments, recently); yields are at a premium & inventory build-up is always an issue; and ROI will scale up more slowly than expected. Not a nice picture, at all. But, these are the conundrums a [comparatively] small company has to go through, once in a while.
That's why I find the co-processor approach interesting: Despite the implementation difficulties (socket; PCB; logic; manufacturers/suppliers/price; etc), it would surely make up for the lack of a new uArch, in the near-term (~2 years), that is, the time Intel claims to have a new one out.
Even after AMD Analyst Day Platform Announcements, I really haven't got a consistent argument to go along with the most optimistic yet unbiased analyst, on what concerns AMD's near to mid-term expectations...
Cheers!